Outcome-Defined Adequacy¶

Prime #: 1044
Origin domain: Philosophy
Subdomain: meta evaluation → Philosophy

Core Idea¶

The adequacy of an artifact, act, or system is specified by whether it produces a defined outcome in its context of use, with the form side deliberately left open — any realization that achieves the outcome counts, and verification is by outcome-measurement rather than form-inspection.

How would you explain it like I'm…

Did It Work?

When you build a pillow fort, what matters is that it keeps the blanket up and you can hide inside — not which exact pillows you used. Outcome-Defined Adequacy means we ask 'Did it DO the job?' instead of 'Is it made the right way?' Any fort that works is a good fort.

Judge By Results

There are two ways to judge if something is good enough. One way checks the steps: did you follow the right recipe and use the right tools? The other way, Outcome-Defined Adequacy, ignores the steps and only checks the result: did it actually do the job in real life? If your paper airplane flies across the room, it's a good plane, no matter how you folded it. The rule is written about the result you want, and any way of getting there counts as long as it works.

Adequacy By Outcome

Outcome-Defined Adequacy is a way of deciding whether something is good enough by checking whether it produces a defined result in actual use, not by inspecting its form or whether it followed the right process. The form is deliberately left open: many different designs might all reach the same result, and only the result is load-bearing. You judge success relationally, against a real-world purpose, instead of against a checklist of features or steps. This contrasts with form-defined adequacy (does it have the right parts?) and process-defined adequacy (did it follow the right procedure?). It comes with its own dangers, too: people can game the measured outcome, cause side-effects, or pick outcomes that are hard to measure honestly.

Outcome-Defined Adequacy is the meta-evaluative stance in which an artifact's, act's, or system's adequacy is specified by whether it produces a defined outcome in its context of use — not by its internal form, surface properties, or process compliance. The form-side is deliberately left open; success is judged relationally, against a real-world purpose or downstream consequence, rather than against an inventory of features or a process checklist. The essential commitment is that the spec or contract is written in outcome terms, not form terms, and any form that reliably produces the outcome counts. The arrangement has recurring roles: an artifact being evaluated, a context of use, a defined outcome, an operational (measurable, observable) specification of that outcome, and form-side openness. Verification is by outcome-measurement rather than form-inspection, and the stance exploits the many-to-one relation between forms and outcomes — many implementations realize the same outcome, only the outcome is load-bearing. It carries characteristic failure modes: gaming, side-effects, measurement difficulty, accountability opacity. Crucially, it sits inside a meta-choice among outcome-defined, form-defined, and process-defined framings, each with its own trade-offs — and the prime's distinctive content is that this framing choice is itself substantive, not a neutral preliminary to evaluation.

Broad Use¶

Philosophy of mind: functionalism defines a mental state by its causal role, not its material realization.
Speech-act theory: communicative success is felicity — illocutionary uptake and effect, not lexical form.
Regulation: outcome-based rules specify an emission or survival target and leave the means open.
Software engineering: black-box testing judges a module by whether outputs satisfy the contract.
Product design: job-to-be-done treats an artifact as adequate if the user accomplishes the task.
Biology: functional morphology and adaptive radiation recognize kinds by ecological role, not ancestry.
Law: performance-based standards and the negligence reasonable-person test are outcome-defined.

Clarity¶

Converts vague "results-orientation" into a precise claim about where the spec is written — outcome side, form side, or process side — and makes the characteristic failure modes of each framing comparable.

Manages Complexity¶

Form-side openness absorbs variety the evaluator would otherwise have to enumerate: an outcome spec admits all valid forms at once, trading enumeration cost for harder verification and harder accountability.

Abstract Reasoning¶

The framing choice is itself substantive — the same artifact can be adequate under one framing and inadequate under another, so adequacy disputes often turn out to be disputes about which framing is in force.

Knowledge Transfer¶

Philosophy of mind → regulation: a regulator weighing outcome-based versus prescriptive standards can borrow the functionalism debate.
Regulation → software: an architect weighing behavioural versus implementation specs borrows from outcome-based regulation.
Universal failure mode: gaming, side-effects, and hard verification carry across as a check on uncritical outcome-framing.

Example¶

An outcome-based emissions rule writes "keep stack emissions below this measured threshold" and leaves the form — scrubber, fuel switch, process redesign — open; verification is by metering the stack, with gaming and pollutant-shifting as the predictable failure modes.

Not to Be Confused With¶

Outcome-Defined Adequacy is not Validation because it is the prior framing choice locating the spec on the outcome side, whereas validation is the verification act that follows — one can validate against a form-defined spec instead.
Outcome-Defined Adequacy is not Goodhart's Law because it is the legitimate stance that invites the hazard, whereas Goodhart is the failure mode in which the measured proxy decouples from the outcome under pressure.
Outcome-Defined Adequacy is not Summative Assessment because it is a criterion-type (outcome with form open), whereas a summative assessment is an occasion-type whose verdict may itself be form-, process-, or outcome-defined.